I used the curve to define a swept line (outward) 1/8" long; used that edge to sweep a new line (45 degrees, 1/8" long)... and used that edge for a 3rd sweep. The graph shows all 3 edges.
When I do this with a curve using a pure cosine equation, all 3 are very smooth.
- r=mean_diam/2
- theta=t*360*turns
- z=(height/2)*cos(t*180)
..but the curve from the Wrap feature would not even recognize the edge of a swept edge for analysis because all the semi-arc tangencies were corrupt.
I would love to see PTC engineering weigh in issues like this.
Again, the severity of this issue doesn't show up with arcs or lines wrapped on a cylinder, only an equation curve becomes corrupted to this extent. But an equation curve itself is near perfect!